Worked on the kedro-org/kedro repository to enhance the robustness and reliability of the datasets API. Focused on backend development using Python, the work addressed a critical bug by broadening the type hints for the pipelines parameter in the list_datasets method. This change allowed the API to accept a union of list[str], list[Pipeline], or None, improving input validation and reducing test flakiness. By strengthening type hinting practices, the update stabilized end-to-end tests and clarified API usage for downstream users. The approach emphasized maintainability and test coverage, ensuring more flexible and accurate handling of datasets within the Kedro framework.